It is well known that large volcanic eruptions contribute to climate variability. However, quantifying these contributions has proven challenging due to inconsistencies in both historic atmospheric data observed in ice cores and corresponding temperature variations seen in climate proxies such as tree rings. Published today in the journal Nature, a new study led by scientists from the Desert Research Institute (DRI) and collaborating international institutions, resolves these inconsistencies with a new reconstruction of the timing and associated radiative forcing of nearly 300 individual volcanic eruptions extending as far back as the early Roman period. 

"Using new records we are able to show that large volcanic eruptions in the tropics and high latitudes were the dominant drivers of climate variability, responsible for numerous and widespread summer cooling extremes over the past 2,500 years," said the study's lead author Michael Sigl, Ph.D., an assistant research professor at DRI and postdoctoral fellow with the Paul Scherrer Institute in Switzerland. 

"These cooler temperatures were caused by large amounts of volcanic sulfate particles injected into the upper atmosphere," Sigl added, "shielding the Earth's surface from incoming solar radiation." 

The study shows that 15 of the 16 coldest summers recorded between 500 BC and 1,000 AD followed large volcanic eruptions - with four of the coldest occurring shortly after the largest volcanic events found in record. 

This new reconstruction is derived from more than 20 individual ice cores extracted from ice sheets in Greenland and Antarctica and analyzed for volcanic sulfate primarily using DRI's state-of-the-art, ultra-trace chemical ice-core analytical system. 

These ice-core records provide a year-by-year history of atmospheric sulfate levels through time. Additional measurements including other chemical parameters were made at collaborating institutions. 

"We used a new method for producing the timescale," explained Mai Winstrup, Ph.D., a postdoctoral researcher at the University of Washington, Seattle. "Previously, this has been done by hand, but we used a statistical algorithm instead. Together with the state-of-the-art ice core chemistry measurements, this resulted in a more accurate dating of the ice cores."